Oninit cdi / Archive
Talk to an expert →

query using cpu

Posted in 2006

Topics: SQL Development & Query Writing

Floyd Wellershaus — — source: Usenet: comp.databases.informix

Any idea why a query that does a join of a table on itself, doing a count and a group by, would cause the cpu usage to go drastically up, almost instantly ? Not pdq, just a query, on a non fragged table. ======================== -<<Floyd Wellershaus>>- Database Administrator Unix Administrator email: fwellers@yahoo.com Home: 703-430-0805 Cell: 703-477-6045 ======================== http://www.one.org/

↪ replying to Floyd Wellershaus
david@smooth1.co.uk — — source: Usenet: comp.databases.informix

Floyd Wellershaus wrote: > Any idea why a query that does a join of a table on itself, doing a count and a group by, would cause the cpu usage to go drastically up, almost instantly ? Not pdq, just a query, on a non fragged table. > > It is doing multiple sequential scans of the table. All the table is cached in the buffers hence no disk i/o hence cpu usage flies up. 1. Query sysmaster:syssesprof to find the session causing the problem and watch its bufreads count flying up! OR 2. onstat -g ses , get the thread id for the thread and check the output from onstat -g tpf for the thread !!

↪ replying to david@smooth1.co.uk
Floyd Wellershaus — — source: Usenet: comp.databases.informix

That's exactly it. I knew I was missing something. Thank for the response !! ======================== -<<Floyd Wellershaus>>- Database Administrator Unix Administrator email: fwellers@yahoo.com Home: 703-430-0805 Cell: 703-477-6045 ======================== http://www.one.org/ ----- Original Message ---- From: david@smooth1.co.uk To: informix-list@iiug.org Sent: Friday, August 11, 2006 4:59:57 PM Subject: Re: query using cpu Floyd Wellershaus wrote: > Any idea why a query that does a join of a table on itself, doing a count and a group by, would cause the cpu usage to go drastically up, almost instantly ? Not pdq, just a query, on a non fragged table. > > It is doing multiple sequential scans of the table. All the table is cached in the buffers hence no disk i/o hence cpu usage flies up. 1. Query sysmaster:syssesprof to find the session causing the problem and watch its bufreads count flying up! OR 2. onstat -g ses , get the thread id for the thread and check the output from onstat -g tpf for the thread !! _______________________________________________ Informix-list mailing list Informix-list@iiug.org http://www.iiug.org/mailman/listinfo/informix-list

Related threads